Pre-Finetuning for Few-Shot Emotional Speech Recognition

被引:1
|
作者
Chen, Maximillian [1 ]
Yu, Zhou [1 ]
机构
[1] Columbia Univ, New York, NY 10027 USA
来源
INTERSPEECH 2023 | 2023年
关键词
emotion recognition; low-resource learning; pre-finetuning; transfer learning; CORPUS;
D O I
10.21437/Interspeech.2023-136
中图分类号
O42 [声学];
学科分类号
070206 ; 082403 ;
摘要
Speech models have long been known to overfit individual speakers for many classification tasks. This leads to poor generalization in settings where the speakers are out-of-domain or out-of-distribution, as is common in production environments. We view speaker adaptation as a few-shot learning problem and propose investigating transfer learning approaches inspired by recent success with pre-trained models in natural language tasks. We propose pre-finetuning speech models on difficult tasks to distill knowledge into few-shot downstream classification objectives. We pre-finetune Wav2Vec2.0 on every permutation of four multiclass emotional speech recognition corpora and evaluate our pre-finetuned models through 33,600 few-shot fine-tuning trials on the Emotional Speech Dataset.
引用
收藏
页码:3602 / 3606
页数:5
相关论文
共 50 条
  • [31] FEW-NERD: A Few-shot Named Entity Recognition Dataset
    Ding, Ning
    Xu, Guangwei
    Chen, Yulin
    Wang, Xiaobin
    Han, Xu
    Xie, Pengjun
    Zheng, Hai-Tao
    Liu, Zhiyuan
    59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING (ACL-IJCNLP 2021), VOL 1, 2021, : 3198 - 3213
  • [32] Cross-Corpus Speech Emotion Recognition Based on Few-Shot Learning and Domain Adaptation
    Ahn, Youngdo
    Lee, Sung Joo
    Shin, Jong Won
    IEEE SIGNAL PROCESSING LETTERS, 2021, 28 : 1190 - 1194
  • [33] Loss Architecture Search for Few-Shot Object Recognition
    Yue, Jun
    Miao, Zelang
    He, Yueguang
    Du, Nianchun
    COMPLEXITY, 2020, 2020
  • [34] Compositional Few-Shot Recognition with Primitive Discovery and Enhancing
    Zou, Yixiong
    Zhang, Shanghang
    Chen, Ke
    Tian, Yonghong
    Wang, Yaowei
    Moura, Jose M. F.
    MM '20: PROCEEDINGS OF THE 28TH ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, 2020, : 156 - 164
  • [35] Shaping Visual Representations With Attributes for Few-Shot Recognition
    Chen, Haoxing
    Li, Huaxiong
    Li, Yaohui
    Chen, Chunlin
    IEEE SIGNAL PROCESSING LETTERS, 2022, 29 : 1397 - 1401
  • [36] Compound Prototype Matching for Few-Shot Action Recognition
    Huang, Yifei
    Yang, Lijin
    Sato, Yoichi
    COMPUTER VISION - ECCV 2022, PT IV, 2022, 13664 : 351 - 368
  • [37] On the Importance of Spatial Relations for Few-shot Action Recognition
    Zhang, Yilun
    Fu, Yuqian
    Ma, Xingjun
    Qi, Lizhe
    Chen, Jingjing
    Wu, Zuxuan
    Jiang, Yu-Gang
    PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA, MM 2023, 2023, : 2243 - 2251
  • [38] Task Adaptive Modeling for Few-shot Action Recognition
    Wang, Jiayi
    Jin, Yi
    Feng, Songhe
    Li, Yidong
    2022 IEEE 24TH INTERNATIONAL WORKSHOP ON MULTIMEDIA SIGNAL PROCESSING (MMSP), 2022,
  • [39] Efficient Few-Shot Incremental Training for Landmark Recognition
    Neuschmied, Helmut
    Bailer, Werner
    PROCEEDINGS OF THE 2024 ACM INTERNATIONAL CONFERENCE ON INTERACTIVE MEDIA EXPERIENCES WORKSHOPS, IMXW 2024, 2024, : 44 - 51
  • [40] Few-shot semantic segmentation for industrial defect recognition
    Shi, Xiangwen
    Zhang, Shaobing
    Cheng, Miao
    He, Lian
    Tang, Xianghong
    Cui, Zhe
    COMPUTERS IN INDUSTRY, 2023, 148